Oppo Digital - DV-970HD

MPEG Maker: Mediatek MPEG Model: MT1389EE Deinterlacer Maker: Mediatek Deinterlacer Model: MT1389EE     MSRP: $149.00 Website: http://www.oppodigital.com

Oppo Digital made a big name for themselves just awhile back with their first entry into the DVD player market: the 971H. That budget player was one of the first to combine a Mediatek MPEG decoder with the popular Faroudja FLI-2310 de-interlacing/scaling chip. What resulted was a player that had an outstanding core video section, excellent de-interlacing, and a great price. But the 971H had its issues. For one thing, it would only do de-interlacing and scaling via its DVI output, leaving component video users in the cold. The Faroudja-based video processing also has issues with some displays causing “macroblocking” artifacts that can make the playback of DVD near unwatchable depending on how bad the interaction between the display and player is. The DVI output is also an issue for some. DVI was originally intended for PC applications, so it is a bit limited in its overall capability. But for those who looked past these issues, the 971H was a steal. Now, Oppo Digital has released an even bigger steal, the DV-970HD. This new player incorporates even more features and flexibility, at an even lower price!

The 970HD is based on the popular Mediatek (MTK) MPEG decoder, just like its predecessor, but the implementation is different this time around. Oppo Digital has eliminated the Faroudja processing and is relying solely on the processing inside the MTK chip. While this does result in a hit in overall video de-interlacing performance, it eliminates the risk of macroblocking artifacts if your display is susceptible to those issues.

The changes don’t stop there. The 970HD doesn’t use a DVI output, but rather the new HDMI standard. This means you can output digital video AND audio. It also means you have choices for output color space and support for high-resolution audio via a single connector. And because HMDI supports 480i video signals (DVI is limited to 480p and up), the 970HD comes in as a very attractive digital transport for feeding an outboard video processor. The 970HD also adds support for progressive playback via the component outputs and SACD support, making it a true universal DVD player. This is without a doubt the most complete feature package I’ve seen from a player at $150. Universal players are normally twice that price at the low end.

Like the 971HD, there is a lot of support for extra features in the 970HD. On the video side, you’ll find support for DIVX, MPEG4, DVD-R, DVD-RW, and DVD Video. There is also smartcard and USB support on the front panel. On the audio side you’ll find support for MP3 and Windows Media Audio, along with both high-resolution audio formats: SACD and DVD Audio. The 970HD supports the new HDMI v1.1 specification, so DVD Audio’s high-resolution multi-channel PCM sound can be output digitally via HDMI. This is also the only player I’m aware of that allows SACD’s DSD sound format to be converted to high-resolution PCM (88.1kHz/24bit) and transmitted via HDMI as well! End users should be aware that your receivers or surround processors need to be HDMI v1.1 compliant as well to use this feature.

From a usability standpoint, the 970HD is about the same as the 971H. Oppo packaged the same remote control that was later offered to 971H owners. While it is still not the best remote out there, it is A LOT better than their old design and gets the job done. The player’s set-up menus are very easy to navigate and have options for setting up the picture and preferences for audio output, including adjustments for audio output levels, crossover, and delay. The 970HD lets you pick between bitstream and PCM for the HDMI output. Unfortunately if you pick PCM, all Dolby Digital and DTS soundtracks will be converted to PCM for output. I’ve never tested a player that has done this before. You can work around this by using the standard digital output for DVD movie soundtracks and the HDMI for high-resolution audio if you intend to use this player for DVD Audio or SACD.

A bug that I ran into with the 970HD is playback of DVD Audio and SACD via HDMI. I am currently using the Anthem Statement D2 A/V processor in my system, and it is HDMI 1.1 compliant. But for some reason, the only way to get the full output support of DVD Audio and SACD is if I set the output of the 970HD to 1080i. Otherwise it is truncated in both resolution and number of channels present. Anything other than 1080i results in stereo only playback, even if 5.1 sound was selected. I informed Oppo and Anthem of the issue, and it is being looked into. Oppo has tested this feature on numerous receivers that support HDMI 1.1 and apparently some have this issue and some don’t, so if this is an important feature to you, make sure you check it out if you purchase this player. Oppo’s customer support is superb, so I am quite sure that they will do anything and everything they can to help.

The speed of the 970HD is outstanding. Disc loads are extremely fast, chapter skips are almost instantaneous, and the layer change is completely seamless. Absolutely no complaints in this department!

I was excited to see how the 970HD would test in our benchmark. I realized right up front that the MTK MPEG decoder would only do so well in our de-interlacing tests, but I was looking at this player more from the transport standpoint, so its core video performance was the most important aspect to me. And Oppo hit a home run here.

When I first received the 970HD, I immediately tested it with an output resolution of 480i. This is the resolution I was the most interested in as I have several outboard video processors at my disposal and a few of them accept 480i via HDMI. Immediately I found a problem. The Mediatek had severe CUE issues when outputting 480i. This was strange, because I’ve tested lots of MTK based DVD players before, and chroma performance is usually near perfect. When I switched the player to 480p, all of the CUE issues went away! Now that is REALLY bizarre. The player did the correct upsampling with a progressive output, but not interlaced. I talked to Oppo about it, and they relayed the information to Mediatek to see if it could be resolved. A few weeks later I got news from Oppo that a firmware fix had been developed! After installing the firmware (software download), I re-checked and sure enough the problem was completely solved! Just another reason Oppo Digital has continued to impress me over and over again with their product support!

The rest of the core video performance testing went quite well. The only limitation I found was the selection of output color space. DVD is mastered in YCbCr 4:2:0, and most (if not all) MPEG decoders upsample this to YCbCr 4:2:2 for output. Unfortunately, the engineers at Mediatek have noticed issues with the 4:2:2 output, so selections for this chip, and the 970HD, are limited to RGB and YCbCr 4:4:4. While not a huge deal by any means, I still like to see support for 4:2:2. Both color space outputs are mapped correctly and pass below black and above white information.

For pixel cropping, I measured all output resolutions for both the HDMI output and the component output. The HDMI output had different results with every resolution. 480i had 1 pixel cropped from the bottom, 480p had no cropping, 720p had 2 pixels cropped from the right side, and 1080i had only 1 pixel cropped from the right side. While we like to see no pixel cropping at all, less than 2 on any side and 2 or less overall for any output resolution isn’t too shabby. The component output was a different matter. The 480i output was the same as the HDMI 480i output, but the 480p output was cropping 4 pixels from the top, 1 from the right, and 3 from the left. Definitely a bit more excessive on this output, and our suggestion is to use the HDMI output if possible.

Another issue I found when I initially got the player was the default white level for the component outputs. It was on the high side, but lowering the player’s contrast setting to -3 fixed the issue. Oppo came through again with new firmware which adjusted the output level to the appropriate level by default. The component output measured in at exactly 100 IRE as it should.

Neither the HDMI output nor the component output suffers from Y/C delay. I measured the component output using a bowtie pattern and a digital oscilloscope, and Y/C timing was perfect. For HDMI, I used a Y/C delay test pattern from the new Anchor Bay Technologies VRS test disc.

The Mediatek MPEG decoder did as I expected with our de-interlacing tests. This chip is slightly better than average, but it can be a bit sluggish when it comes to locking on to a cadence. It does trip up with a few of our more difficult tests, but I expected as much from an all-in-one chip solution. What did surprise me though was the Mediatek’s diagonal line processing. Using the “3 angles” test from the HQV Benchmark DVD, I was pleased to see that Mediatek has developed some type of diagonal line processing for video-based material. While it wasn’t quite to the level of popular solutions like the Realta HQV or Faroudja DCDi, it did a very commendable job with the lines.

The Mediatek chip did a fine job with the majority of our 3-2 based cadence patterns, including incorrectly flagged ones, video flagged, and high detail. It also supports a 2-2 cadence pretty well with no obvious artifacts in our test using the Natural Splendors DVD.

Overall, I would say the Mediatek did an above-average job in our de-interlacing tests for an all-in-one solution, but it still does not live up to the bar set by the higher line video processing chips out there now. Keep in mind though, this player costs a fraction of the price of players normally incorporating those chips.

Conclusions

Once again, Oppo Digital has proved that you don’t need to spend a lot of money to get an outstanding product. The 970HD would make an excellent digital transport for feeding a standalone video processor, but it also holds its own very well too. Support for both DVD Audio and SACD via HDMI is just icing on the cake. If you are looking for a solid transport, then add the DV-970HD to your short list of capable players. If you are looking for a great budget DVD player, this may just be your ticket.

Video Frequency Response
Oppo Digital

Toshiba - SD-K860

MPEG Maker: Zoran MPEG Model: Vaddis 888 Deinterlacer Maker: Zoran Deinterlacer Model: Vaddis 888     MSRP: $79.99 Website: http://www.toshiba.com

The SD-K860 is Toshiba’s newest budget entry into the digital DVD player category. It is the least expensive HDMI based DVD player I’ve seen from Toshiba to date, and it offers commensurate performance to its price point. It is exciting to see HDMI based DVD players coming in under $100, but so far my experience with them hasn’t been very good. For our Benchmark, two big manufacturers have really let me down with their offerings and it all comes down to the core performance of the player, which is crucial. The SD-K860 definitely has some issues in this department.

The K860 is based on Zoran’s newest all-in-one MPEG decoder/video processing chip, the Vaddis 888. I have evaluated a lot of players that use Zoran chips and their performance is never consistent between models for some reason, so I was interested in seeing how this new chip would fare compared to their older models.

Immediately I was impressed with the chip’s chroma performance in our CUE tests. The last few Zoran chips have done well here, and this one seems to improve even more on their performance. The K860 passed all of our CUE tests, including the 4:2:0 ICP test, which is VERY rare. This MPEG decoder also handles layer changes very well with a near seamless layer change delivered with our difficult full bitrate test.

The MPEG decoder only outputs 4:4:4 YCbCr with no other selections available. The HDMI output supports 480p, 720p, and 1080i, and the component output supports 480i and 480p. There are quite a few differences in terms of performance between the two outputs, with the component output fairing a bit better in our core video tests.

The HDMI output suffers from over a full pixel of Y/C delay. This lends to a softer picture, as the two fields of the image are not converged properly, so there are some slight outlines on objects. The HDMI output is also clipping head and toe room. Video is encoded with black and white at digital 16 and 235 respectfully in an 8 bit system (255 levels). This player is truncating the information that falls after those points. Some don’t see this as a big deal, but it does not preserve the intended dynamic range of DVD, and it also creates issues for calibration.

There is absolutely no pixel cropping via HDMI at any resolution. This was nice, and a rarity in player testing lately.

The component output retains the full dynamic range of video and does not clip below black or above white information. The white level measured in at a respectable 99 IRE, which is close to the reference 100 IRE it should be. The analog video frequency response was also quite flat, with no artificial ringing or smearing in the upper frequencies. There is some slight pixel cropping via component, with a combined 5 pixels being clipped from the sides of the image, and 2 pixels cropped from the bottom. I don’t know why the component output has pixel cropping, but the HDMI output does not. They should match each other.

From a usability standpoint, the K860 could use some work. Load times - when you insert a disc - are a bit sluggish, and operability - once a disc is loaded - is slower than I would expect from a later generation DVD player. Menu navigation was tedious at times, with a lot of lagging in between selections. Chapter skips were pretty fast though, once a feature was started. The overall user interface and set-up was easy, if not a bit limited.

Moving on to our de-interlacing tests, the Zoran chip did pretty well for this price point. The player passed the majority of our tests, with the exception of our Bad Flagging tests. The player failed our 3-2 Mixed Flags and 3-2 Video Flags tests along with our Incorrect Progressive Flags test. So the player isn’t really cadence based, though it did perform cadence detections in some of our other tests. The K860 does handle a 2-2 cadence for video based material but its diagonal line processing is terrible, with severe jaggies using the HQV test disc material.

Conclusions

Overall, I found the SD-K860 to be a sub-par DVD player. There were some highlights, but the player needs work in its core video performance. This has become an unsettling trend lately in our tests. While I don’t expect top of the line performance at this price point, the issues found here should just not be a problem regardless of the price. Limitations in video processing are expected, not limitations in core performance. While this is an attractive price point for a HDMI based DVD player, I would have to recommend looking elsewhere.

Video Frequency Response
Toshiba

Pioneer - DV-393-S

MPEG Maker: Mediatek MPEG Model: MT1389FE Deinterlacer Maker: Mediatek Deinterlacer Model: MT1389FE     MSRP: $79.00 Website: http://www.pioneerelectronics.com

The DV-393 (the -S at the end of the model number just means it is the silver version vs. the black version) is the less expensive sibling to Pioneer’s new budget HDMI player, the DV-490V. Everything is exactly the same between these two players with the exception of the HDMI output on the 490V. Since the price difference is about $20, I would have to recommend the 490V. While your display may not have an HDMI input right now, later down the line, you will buy a display with HDMI, because they all will have them, and you will probably want to go the digital route at that time.

The DV-393 is one of the least expensive Pioneer DVD players I’ve reviewed to date. It is nice to see some of the more established manufacturers release products at this price point, but it seems like this model has a lot of issues compared to Pioneer’s normal offerings. This was also the case with the DV-490V.

The biggest issue with this player is its core video performance. There is just way too much amiss here, and this is the most important aspect of a DVD player’s performance. Consider it the framework. You can have the best video processing and de-interlacing on the market, but if the framework around it doesn't perform well, then it doesn’t matter.

The DV-393 has a lot of issues in this regard. First off, the player is clipping an excessive amount of pixels from every side of the image. This means part of the image is literally not getting to the end display. We measured over 10 pixels on any one side, which is way outside of the normal range we see with most DVD players. The next problem is Y/C delay. NTSC video is comprised of Luma (black and white) and Chroma (color) information overlaid on each other to create the picture you watch on the screen. These two separate signals need to be superimposed on top of each other perfectly. If it doesn't, then you essentially get convergence issues, similar to what you find with older CRT displays. This creates a softer image, plus outlining on one side of objects.

Next up was the white level of the player. A pure white signal should measure out at 100 IRE. For this test we use a grayscale ramp and a digital oscilloscope. The white output for this player measured in at 97 IRE, which is just below our passing criteria.

The DV-393 did do well with some of our other tests though. The player does pass a below black pluge, and its chroma performance is very commendable. The Mediatek MPEG decoder used always does very well in this department, and the 393 sailed through our CUE tests, failing only our 4:2:0 ICP test. The Mediatek chip cannot keep the CUE issue at bay with motion during this test, and you will see stripping with some material. The video frequency response of this player is also quite good and nearly ruler flat.

The component output only supports 480i and 480p like most analog-based progressive DVD players. All of my tests were done at 480p. The Mediatek chip’s performance was exactly the same as its higher priced sibling, the 490V. The chip passed all of our tests with the exception of the 3-2 Mixed Flags and 3-2 Video tests. It also failed our Bad Edit test with some slight combing using both our Big Lebowski clip and a test clip from Faroudja.

From a usability standpoint, this player is quite good. The Mediatek chip is known for its speed in navigation as well as its quick load times, and the 393 didn’t disappoint. Menu navigation and chapter skips were zippy, with no hitches to speak of. The layer change was not quite seamless, but with only the briefest hiccup using our high bit rate scroll test. Just blink and you’ll miss it!!

Conclusions

The DV-393 is an average to below-average progressive scan DVD player. At this price point, I wasn’t expecting much, but with just a bit more attention to detail, this could have been a really great value overall. Too many DVD player manufacturers are losing sight of what makes a quality picture and keep messing up the core performance that is so important. The DV-393 is no exception. While I don’t think people can expect much at this price point, I think they should expect better than this. In fact, they should insist on it.

Video Frequency Response
Pioneer

Panasonic - DVD-S52

MPEG Maker: MEI MPEG Model: Unknown Deinterlacer Maker: Unknown Deinterlacer Model: Unknown     MSRP: $99.95 Website: http://www.panasonic.com

The DVD-S52 is one of the least expensive HDMI-based DVD players I’ve tested to date. At this price you wouldn’t expect much in terms of features, but Panasonic delivers a decent package of options here for the budget minded consumer. Not only does it offer an HDMI output, but this output is HDMI v1.1 compliant, and the player supports full playback of DVD-Audio. This means you can use one cable to deliver both standard DVD video and audio plus high-resolution PCM audio from DVD-Audio software. The player also supports upscaling of DVD via HDMI to 720p and 1080i. While this isn’t the first time Panasonic has offered these features to the consumer market, it has never been quite this inexpensive.

The S52 is based on a Matsushita (Panasonic) MPEG decoder. I did not see any other video processing chips when I opened the player up, and the video performance doesn’t suggest that Panasonic is using any. In the past we’ve been big fans of Panasonic’s MPEG decoders. In fact, the Panasonic MPEG decoders found in the RP-91 and RP-82 continue to be our standard to this day. Those players are over 5 years old now! This new chip does not live up to that same standard, unfortunately. The implementation on the digital side is good, but the analog output of the chip needs quite a bit of work.

I started my tests with the HDMI output. I fed a Panasonic AE-900 LCD projector directly and also put a DVDO VP30 in between for some of the tests. The VP30 allows me to verify the output color space of the player plus do some more extensive CUE testing.

The core performance on the HDMI side was quite good. The S52 is not cropping any pixels via HDMI and retains the full dynamic range of the luma signal. It does not clip below black or above white information in the image. The player also doesn’t have any signs of Y/C delay via HDMI, which helps with the overall sharpness of the image.

The S52 does suffer from some slight flickering in our CUE testing with alternating 3-2 cadence based material. This cadence is common in some of the early Pixar and Disney DVDs and can be noticeable at times. All of the other CUE tests were passed, including the 4:2:0 ICP test which means this player does some chroma filtering.

The core performance of the component (analog) outputs was completely different and not nearly as good. There is some mild pixel cropping with 2 pixels missing from the left side and 3 from the right, and there is over a full pixel of Y/C delay. All NTSC video is transmitted with both black and white (luma) and color information (chroma) placed on top of each other. They are separate and distinct signals, and if they are not overlaid exactly on top of each other, you will see outlining or “ghosting” around the image on one side. This player delays the chroma, so you will probably see a faint outline on the right side of objects.

The overall white level of the analog output was also extremely low. Using a horizontal gray ramp and a digital oscilloscope I measured the full IRE sweep of the player from 0 to 100 IRE. If the player’s output were correct, the respective 0 and 100 IRE points would measure as such on the scope. In this case the 100 IRE point measured REALLY low and came in at 91 IRE.

The analog frequency response of the player was good, with a fairly flat curve throughout the full video bandwidth.

From a usability standpoint, the Panasonic was about average. Menu navigation and chapter skipping were quick enough and never distracting. Disc load times were also good. The layer change was a tad sluggish at 1.5 seconds, but livable given the price point. I wasn’t that impressed with the advanced setup menus. Settings for HDMI output and de-interlacing mode were a bit hidden, so some users may have difficulty finding them. The normal setup menus were easy to use though.

I would have liked to see more attention to detail on the analog side of this player, especially since the HDMI (digital) output was so good.

Moving on to our de-interlacing tests, the S52 starts to show the limitations of one-chip mass-market MPEG decoders. The S52 offers three separate de-interlacing modes; Auto 1, Auto 2, and Video. Video is a forced video mode that will not lock onto any cadences or switch into a film mode. This results in a much softer image than the Auto modes but will result in very little, if any, combing. The only real difference between Auto 1 and Auto 2 was the latter’s ability to lock on to a 2-2 cadence correctly. This is only important though if you know you are going to watch material mastered with this cadence. Some TV based DVDs use this. Otherwise, we would suggest just sticking with Auto 1 (default), as it handles most of the film-based cadences.

The S52 does not handle breaks in cadence or mixed cadences very well at all. It is also not motion-adaptive, so it will not handle video-based material very well. This pretty much defines the bottom end of video de-interlacing in mass-market progressive scan players. Even our medium difficulty tests tripped this player up, resulting in drops to video mode and loss of overall resolution. It did perform nicely at recovery from video to film mode as cadences changed. At this price point I wasn’t expecting much better, but I hoped that Panasonic delivered performance closer to their players that cost just a little bit more.

Conclusions

It would be hard for me to recommend this player to anyone. The video processing is on the weak side and will result in a softer image with poorly mastered material. I was impressed with the feature set considering the price point, but that doesn’t make up for the other issues. For those using component inputs, you should steer clear of this player, as its core performance using that output needs a lot of work and there is a lot better to be found in this price range.

Video Frequency Response
Panasonic