Tuesday, August 18, 2009

Lossy vs. lossless results

12:00 AM / Blu-ray / Comments4 Comments

Blu-ray

On Sunday, I wrote a post about the issues surrounding the use of lossy screen captures of Blu-ray Discs, arguing that there was no real world benefit to using a more space- and bandwidth-heavy lossless format such as .BMP or .PNG rather than lossy JPEGs. I also posted a selection of images and challenged readers to identify which were lossy and which were lossless, promising that I would provide the results on Tuesday.

In case you didn't see the captures last time round, here they are again:

 
#1. Monsters, Inc.

Monsters, Inc.
Image 1 / Image 2

 
#2. Franklyn

Franklyn
Image 1 / Image 2

 
#3. The Dark Knight

The Dark Knight
Image 1 / Image 2

And now for the results:

  • For both Example #1 and 2, Monsters, Inc. and Franklyn respectively, the lossy JPG is Image 2.
     
  • Example #3, The Dark Knight, was a trick question: the actual picture area is identical, neither of them having been saved lossily. In case anyone decided to look at the file sizes and thereby work out that they were both the same, I played around with the black letterbox area on one image, removing some compression guff that was present in the original image, thereby altering its file size.

A few people posted their responses publicly, and a couple more emailed me theirs. Interestingly, no-one got all of them right. One person, who emailed me his answers, got Examples #1 and 2 correct (albeit with the proviso that he was only guessing), but no-one worked out that the two captures from The Dark Knight were the same. The last one was, you might think, a bit sneaky of me, but I never did say I was going to play fair. In any event, I did it to make a serious point: tell people that two images were saved differently, and many will begin to think they see differences, even where there aren't any.

Well, there you have it. I don't expect this to put the debate to bed any time soon, but hopefully it provides something of an explanation as to why I not only use JPEGs but also see no problem with doing so.

 
4 Comments

1. Marc said:

Hi Michael,

I don't mind JPEGs one bit. When I'm reading reviews, I like to look at the screenshots, but I never STUDY them. JPEGs are fine for giving you an impression of the image quality.

Now, I can't seem to find an email address of yours and I wanted to ask you a question, so I'll just do it here. A while ago I watched an interlaced Blu-ray, and I could not find any evidence watching the movie closely that it was really interlaced. And having just read DVDBeaver's review of Brick on Blu-ray, which they say is interlaced, again looking at the screenshots or reading their observations I don't see the problems you would normally see on an interlaced DVD (like combing). My question to you is, do you have any experience with interlaced Blu-rays, and what did they look like (compared to an interlaced DVD)? PowerDVD doesn't let me turn off their de-interlacing feature, which is either doing a great job of the interlaced movie I'm watching is not interlaced as advertised on the box OR Blu-rays can be de-interlaced much better than DVD's. Or something else ;-)

(Posted on Tuesday, August 18, 2009 at 4:59 PM)

I'm not Michael, but I suspect you were watching the disc on a system that was performing film mode processing. In other words, the interlaced source was being converted losslessly to progressive upon playback, so you will have seen no difference.

Even if the playback device wasn't doing this and was processing in Video mode, the resulting jaggies are so small that many won't ever notice them. At least not with most film content.

Whether or not what DVD Beaver's screen grabs do or don't show depends how they are taking the grabs (don't even get me started on that!). If the grabs are of a 50hz disc (2-2 cadence), then while the video is technically interlaced, both fields contain the exact same data, so it might as well be Progressive. In a case like that, Interlacing is really just a technical concept.

(Posted on Wednesday, August 19, 2009 at 2:59 AM)

3. Author Profile Page Marc said:

Hi David,

Thanks for your insight. Is there any way I can check if my system (a PC with PowerDVD version 7 and 9) is processing in video/film mode?

Also, what is the best way to take screencaps? I use Media Player Classic (with PowerDVD filters processing the video), because that seems to be the easiest way, but I'm not sure if this method 'colours' the image somewhat. I'm also aware of ArcSoft TotalMedia Theatre 3 Platinum, which has a build in screenshot feature (you have to use AnyDVD HD for this to work though), so that would be even easier and faster but when I compare TotalMedia with PowerDVD, the image looks noticeably worse (very soft, for some reason).

(Posted on Wednesday, August 19, 2009 at 2:01 PM)

4. Author Profile Page Michael said:

Hi Marc,

Sorry for the late response. Looks like David's already answered your initial question, but I'll do my best to respond to your follow-up. As far as I can gather, using Media Player Classic is as good a way as any to take screen captures. Personally, I use an AviSynth script which I then load in VirtualDub, which can be a little time-consuming due to how slowly it seeks through the footage, but has served me well so far. I didn't realise ArcSoft had put out a Blu-ray compliant player - I may have to take a look at it.

By the way, my email address is whiggles[at]ntlworld[dot]com. I should probably make that more readily visible on the site.

(Posted on Thursday, August 20, 2009 at 2:44 PM)

 
To combat spam, commenting is automatically disabled on entries older than 30 days.

Did a comment you tried to post accidentally get eaten by the spam filter? It happens from time to time. I get upwards of 200 spam comments every day and unfortunately don't have the time to weed through all of them in case something genuine ended up there by mistake. If one of your posts gets incorrectly flagged as spam, email me at whiggles[at]ntlworld[dot]com and I'll do my best to retrieve it.