Quoting Grendell:

Quote:
Quoting Winston Smith:

Quote:
I don't believe that ajy of you or myself included can claim to be accurate enough to gived credible calculations relative to the AR, and for those of you who have seen the BD GWTW then you know the level of visual detail I refer to and that was only because of 5 measly pixels. o stop with pretending a level of skill which you simply do not possess, it;s pure unmitigated BULL

I am not pretending, anyone can do it to down to the pixel by using the "zoom" function in any image editing program. How can people not be aware of this. It boggles my mind.

  We'll try to help our friend Skip :

Quoting widescreenforever:

Quote:
why?  Because if Warner's releases  i.e. The Dark Knight on standard DVD and BluRay the standard DVD is listed on their package as 2.40:1 and the BD is issued with 2.4:1 on their packaging .. so far so good..

Until you contribute them that is and If you put the same on either package it will get voted down.. Why?
it's as you say .. " Maybe I am missing something but, because those two ratios are exactly the same...2.4=2.40=2.400=2.4000...I don't understand why that would bother anybody."  ...

It seems to bother the voters here .. 

It doesn't just bother the voters, it bothers Invelos - and rightly so. We shouldn't go by the whims of what the various DVD distributors use on their packaging, but we should use a standard format. In our case, that's 2.40:1, not 2.4:1, even if the latter is on the cover. Happily, there's a contribution filter that automatically "fixes" any incorrectly formatted aspect ratios in the 2.4:1 format to the proper 2.40:1 format. Since this is now something that is automatically enforced for the online database, it's something we no longer have to worry about.

Quoting RHo:

Quote:
... specially in compressed images.

I was lazy and took a jpg image that I had already on my computer. When you take native images in png or bmp, you hardly have doubt about where is the limit, even with the compression used by DVD maker. In my example, I have personally no doubt, even in compressed format, the only real difficulty being for very dark images. But I have not yet seen one movie where you can not find images with clear transition zone.

I mostly agree with the majority of what I read on these forums. But, I'm real a _ _ l when it comes to data. If my cover says 2.40:1 that is the AR I use. If my cover reads 2.4:1, that is the AR I use then lock it. As they are the exact same thing, I want my profiles to reflect what is seen on the cover. All that hinges on whether the cover is reporting the actual AR. We know that a lot of discs that report 1.85:1 are actually 1.78:1, I use 1.78:1.

Note: I do round up or down to the nearest hundredth when measuring using JR Ruler. I voted Do something else, since neither of the other choices reflect what I do.

Quoting mreeder50:

Quote:
If my cover says 2.40:1 that is the AR I use. If my cover reads 2.4:1, that is the AR I use then lock it.

I honestly can't wrap my head around that desire, but that's probably just me, so never mind. The key is that it really doesn't matter: as stated before, Invelos automatically fixes 2.4:1 to 2.40:1, so it's not an issue. And indeed: doing things however you like them and then being able to lock the field is one of the strong points of this program. As long as we've got such a contribution filter to ensure some consistency in the online database (more of that, Ken, much more!), it's all good.

Quoting RHo:

Quote:
Quoting surfeur51:

Quote:

I was lazy and took a jpg image that I had already on my computer. When you take native images in png or bmp, you hardly have doubt about where is the limit, even with the compression used by DVD maker.

I have been talking about the MPEG compression on DVD and not the JPEG compression of your image.

I had understood that, that is why I wrote what is bolded.

Quoting RHo:

Quote:
Is the vertical grey or black line part of the image?

I agree that on dark images you may have doubts. That is why we have to make the measurement on bright images, as I wrote :
Quoting surfeur51:

Quote:
...the only real difficulty being for very dark images. But I have not yet seen one movie where you can not find images with clear transition zone.

Surfeur:

If I am not mistaken your marking line appears to be several pixels OFF in your zoomed image. Like I said anyone who believes they can be accurate is only dreaming. I have said this befiore, but it simply isn't going to happen. This is one area that unless I can see an obvious problem, which happens all too often, then the box and industry standards are operational, I am not interested in someone's imaginings about how perfect they THINK they can be in measuring an image because I know better. And with over 25 years in the video industry, I'll take my experience over yours any day of the week and twice on Sunday.

Thank you for all this.  Lots of interesting thoughts and views.

For the record, if you’re short-sighted you can generally get closer to things before you cannot focus on them.  Ask an optician.  I find it quite simple to measure the size of a picture on a screen down to a single row or column of pixels.  You just pick a nice bright shot, pause it and away you go.  I can’t prove it here, but it’s a fact, take it or leave it.  I’m told some people who do a lot of graphic design work on a computer do this, so it presumably works for them too.

However, regardless of the ability of someone to focus this closely and the ever present possibly of him/her making a mistake, there are some practical issues with trying to measure specific DVDs in this way.  The last line of pixels can sometimes be a bit faded (which RHo has given a nice example of above), but so far when I’ve done the relevant calculation using two slightly different figures to take this into account, it hasn’t been that critical in terms of selecting a ‘best-fit’ to the drop-down box options.  There are also some DVDs that can be a bit more of a mess around the edges, but again on the few where this has been an issue, the range of measurement possibilities hasn’t caused a problem in terms of determining which ‘drop-down’ box choice to use.  If it did I’d just go with what’s on the sleeve.  Finally, subtitles can be an issue if they stray onto the black area below the picture and I have to say this is one where it can be difficult to work out where to measure to and where the difference can be significant.  I’m not saying anyone should simply measure the screen and just accept that, but it’s a way of confirming  the aspect ratio information given elsewhere and highlighting where there might be an issue worth further consideration.

There seems to be two issues that have developed in this thread.  One is whether measuring the screen using an on-screen tool (such as JRuler) can actually be done accurately enough for it to be an appropriate way to document contributed data.  The second is when appropriate data is contributed, should it reflect the options offered in the aspect ratio drop-down box or the actual aspect ratio even if this is a non-standard size.

Personally I always measure the picture but then pick the nearest ratio in the drop-down box to what I’ve measured.  I suspect selecting from the latter is probably what most people do, as many of the contributions seem to take the information they contribute from the sleeve on good faith without measuring anything.  There’s only been a couple (as far as I can remember) where the measurement I’ve made hasn’t agreed with what’s on the sleeve (example below).  In a way, putting an accurate measurement into the database would be good, but I’d rather have a database that’s used consistently by everyone and go with that option (as I think I do now), rather than have people contributing information based on different interpretations of what to do and the data ending up as a messy mixture of both.

I’ve been through a few of my recent contributions to see what I’ve said in my notes for these.  (There have been a lot more, but if what I measure agrees with what’s already in the database, I don’t always include this in my notes.)  If anyone wants to have a go and see what they can come up with a different measurement, you’re very welcome.

The Scarlet Tunic (5-060037-261329)
Taken using JRuler as 753x412 = 1.83, which is nearest to 1.85:1
Sleeve says 16:9

La Posesión (8-420018-199282)
Taken using JRuler as 935x573=1.63, which is nearest to 1.66:1
Sleeve says 1.85:1

Sorry, Haters (807839-003048)
Taken using JRuler as 745 x 404 = 1.844, which is nearest to 1.85:1
Sleeve says 1.85:1

Punk & Disorderly (5-013929-920651)
Taken using JRuler as 734x568 = 1.29, which is nearest to 1.33:1
Sleeve gives no aspect ratio information

Happy Easter everyone.  Extra time for watching DVDs.