Does anyone know if the PQ is better [the same? worse?] on the original, more expensive Tivo Series 3?

I am currently comparing recordings on a TivoHD, LG LST3410a, and my computer made from last night's football game. I'm creating a HD DVD of the recording I made on my computer and playing it on my Toshiba HD-A1. It looks like the LG may have the best PQ of the three, but I really need to do a lot more testing. One thing I've notes is that the LG looks good when playing, but terrible when paused.

I have an S3 but not a TiVo HD, so I haven’t been able to make any A/B comparisons. But based on what I have read about this issue in other venues, the consensus seems to be that there is no discernable difference in PQ between the S3 and THD.

Output of the TivoHD and Series3 is identical when using the same resolution and the same connection.

Did you use the same output type? And do you have both boxes set to output the same resolution? Obviously if the LG is set to output 1080i and the Tivo is set to output 720p, the results will not be comparable. Similarly, if one box is using component and the other is using HDMI, the results will not be comparable.

NBC uses 1080i for its broadcasts, so if either box was set to native, hybrid, hybrid 1080i, or fixed 1080i, and you were using HDMI, then you should not see any difference. If you use component or have either box set to 720p, then that would involve extra processing. The 1080i->720p processing could differ between the LG and Tivo boxes, as could the HDMI->component conversion.

gwsat-
Thanks, when I looked at the specs I assumed that the TivoHD would be at least as good as the S3. It seems that because of the increased processing power and memory in the TivoHD that it should be capable of having better PQ than the S3. However, the differences I'm seeing are fairly slight, so it would be nice to know if anyone has actually done an A/B.

bfdtv-
Good points. Unfortunately, my LG LST-3410a does not have a HDMI output. My TivoHD uses 'Native' and I'm using its HDMI output. My LG LST-3410a uses 'Native' and I'm using its Component output. My receiver transcodes component output to HDMI and sends it to my projector.

I assumed that this would mean that the 3410a would have worse PQ. Just so I'm comparing the same things, I'll use the TivoHD's Component output.

I also noticed that the broadcast that I recorded on my computer actually contains most of the bad stuff that I was seeing when watching the TivoHD. It may well be that the reason the TivoHD looks worse is because it is doing the better job of showing what is being broadcast. Using Component cables from the Tivo could actually be the better choice because it may smooth over some of the digitization I saw.

I'll test this afternoon.

Similarly, if one box is using component and the other is using HDMI, the results will not be comparable.
This has not been my experience. I have a circa 2003 HDTV that has DVI that is not HDCP compliant. Thus I have to connect any DVR to it via component.

I had always been curious about whether there was any discernable difference in PQ between HDMI and component so in August 2006 I did an A/B comparison between HDMI and component using my then brand new 32 inch Sony Bravia XBR and an SA 8300HD DVR. The PQ via component was indistinguishable from what I was seeing via HDMI.

In January, immediately after I got my S3, I went into mad scientist mode once again and made the same A/B comparison using my XBR and my new S3. The results were the same.

Based on what I have read about this issue virtually everyone who has run such tests has reached the conclusion I did: there is no difference in PQ between component and HDMI.

This has not been my experience. I have a circa 2003 HDTV that has DVI that is not HDCP compliant. Thus I have to connect any DVR to it via component.

I had always been curious about whether there was any discernable difference in PQ between HDMI and component so in August 2006 I did an A/B comparison between HDMI and component using my then brand new 32 inch Sony Bravia XBR and an SA 8300HD DVR. The PQ via component was indistinguishable from what I was seeing via HDMI.

In January, immediately after I got my S3, I went into mad scientist mode once again and made the same A/B comparison using my XBR and my new S3. The results were the same.Perhaps I should have worded differently.

My point was really that involving extra A/D parts puts more variables into the equation. No two A/D parts from different vendors produce 100% identical output. Moreover, different displays handle HDMI and component inputs differently. Some filtering may be applied with A/D conversion in the display, which may result in slight differences to resolution and/or colors.

With my DLP and plasma, component and HDMI look very different using the same settings. My displays store separate settings for component and HDMI, and if I use the same settings for each, the pictures are not comparable at all. I can get the HDMI and component inputs to look virtually identical, but only after carefully calibrating each input. Component looks slightly softer than HDMI on the DLP, but I can't tell with most content at my usual viewing distance.

Good points. Unfortunately, my LG LST-3410a does not have a HDMI output. My TivoHD uses 'Native' and I'm using its HDMI output. My LG LST-3410a uses 'Native' and I'm using its Component output. My receiver transcodes component output to HDMI and sends it to my projector.What receiver do you have? You aren't using any upconversion in the receiver, are you?

Perhaps I should have worded differently.

My point was really that involving extra A/D parts puts more variables into the equation. No two A/D parts from different vendors produce 100% identical output. Moreover, different displays handle HDMI and component inputs differently. Some filtering may be applied with A/D conversion in the display, which may result in slight differences to resolution and/or colors.

With my DLP and plasma, component and HDMI look very different using the same settings. My displays store separate settings for component and HDMI, and if I use the same settings for each, the pictures are not comparable at all. I can get the HDMI and component inputs to look virtually identical, but only after carefully calibrating each input. Component looks slightly softer than HDMI on the DLP, but I can't tell with most content at my usual viewing distance.
My DLP is the 2003 model that lacks HDMI so I have never been able to make a component v. HDMI comparison on it. My Sony Bravia XBR is LCD, of course, and it’s the only one I could comment about. I don’t have a plasma but I wish I did. Anyway, I have no disagreement with what you posted. I should have known that you had thought it out.

I confess that I was a bit worried early on about having to use component because of the A/D issues you have addressed. So I was pleasantly surprised when I discovered that I did not see a whit of difference between it and a purely digital HDMI hook up. Of course, given the age of my eyeballs, you might want to consider the source.

What receiver do you have? You aren't using any upconversion in the receiver, are you?
I've a Denon 4308. And I'm not doing any upconversion or other type of conversion on HDMI sources. However, the receiver does take component sources and transcoded them to 1080i (if needed) before sending them out via HDMI to my projector. Since the component sources are going through an extra A/D step in my receiver, I expected that signal to be slightly worse than the pure digital path coming my TivoHD. The signals are very similar.

After carefully checking all three methods of recording, I've decided that they are almost all exactly the same. The PQ advantage I thought I saw in the LG 3410a was because I'd not looked at the exact same sequence. This time I went to bad sections I found on the computer recording and checked them on the other two sources. These sections appear to be identical. Then I went to the crisp sections of the broadcast (like the guys talking in the booth at the start of the game), and all three appear to be identical.

So, I have to rethink my first post and say that the TivoHD is doing a fine job of recording. Unfortunately, Sunday Night Football is not doing all that good a job broadcasting.

So, I have to rethink my first post and say that the TivoHD is doing a fine job of recording. Unfortunately, Sunday Night Football is not doing all that good a job broadcasting.The Tivo will tell you the size of a recording if you hit info, which you can multiple by 8192 and then divide by the number of seconds to find the bitrate.

If you can pull in locals from two markets, you might want to compare the average bitrates on all networks to see which is most likely to produce the best picture.

I did this for some programs here in DC/VA/MD:
Network Title Minutes Size(Gb) Bitrate (Mbps)
WRC Tonight Show w/ Leno (9/4) 62 7.51 16.54
WBAL Tonight Show w/ Leno (9/4) 62 7.03 15.48
WTTG House (9/4) 60 4.42 10.06
WBFF House (9/4) 60 4.36 9.92
WJLA Grey's Anatomy (9/6) 61 5.31 11.89
WMAR Grey's Anatomy (9/6) 61 5.65 12.65
WUSA U.S. Open Tennis (9/7) 60 7.24 16.48
WJZ U.S. Open Tennis (9/7) 60 8.01 18.23
DC/VA/MD residents will get the best picture by watching WRC (DC) for NBC content, WMAR (Balt) for ABC content, and WJZ (Balt) for CBS content. WTTG and WBFF are a wash, which isn't surprising given the way FOX distributes programming.

The Tivo will tell you the size of a recording if you hit info, which you can multiple by 8192 and then divide by the number of seconds to find the bitrate.
bfdtv -- Good call! I didn’t know that. Thanks to your how-to post I have built myself an Excel bitrate calculator and am now armed and dangerous. Heretofore, I had believed that everybody who was giving bitrate numbers must have had a fancy meter of some sort. I have been seeing those file size numbers ever since Version 8.3 of the TiVo software rolled out and didn’t know what they could do until now. Thanks!

bfdtv-
Thanks for the description on how to calculate bitrates. I'm fortunate to have the file on my computer. I cut all of the commercials using VideoReDo (VRD) and then have VRD process the file. When VRD is done it tells me the actual bitrate. For example, that football game had had a bitrate of 14.5Mbps. The funny thing is that the header for the file claims it has a 35Mbps bitrate.

Is the bitrate for a 720p station like Fox directly comparable to the bitrate for a 1080i station?

As a follow-up, this afternoon's Patriots game had a bitrate of 17.5 and looked much, much better.