What size is 6x7 in PX

Image sizes

Note: I have already written a detailed, multi-part article about the image dimensions elsewhere, but it is intended more for readers who want to deal intensively with this problem. At this point I would like to present only the most important things briefly and as clearly as possible.

Pixels and only pixels!

I'll put the most important things right at the beginning. When it comes to image editing, only the number of pixels (length and width) is important. So if you specify the size of a digital image as 2400 x 1800 px, then that's it everything said what is important for the image size in the context of image editing. The other two dimensions, which are mentioned again and again and unfortunately come to the fore, are completely irrelevant and only play a marginal role, for example for assessing a picture or as information to the printer. These secondary dimensions are, on the one hand, the metric image size in cm, inch, point etc. The dimension is completely unimportant when editing the image! The other secondary measure is that relative resolution in dpi (not to be confused with the absolute resolution when recording). Also unimportant, superfluous, even dangerous if you get involved.

And of course the file size in MByte is totally unimportant. It depends on the number of pixels, but even more on the quality set, the color depth, the structures in the image, etc. When editing images, it is wise not to pay attention to the file size at all, even if Picasa stupidly displays it all the time. And it would be wise to ignore the image size in cm (metric image size) or the resolution in dpi, even if Photoshop constantly pushes these dimensions into the foreground and even hides the only important dimension, the number of pixels, a little.

Why are the metric image size and resolution unimportant? Quite simply because they say almost nothing, and are even interchangeable without the image being changed in any way. Simple example (you can easily see it in the size settings window of every program):

We take the picture of 2400 x 1800 pixels again. If we open the size dialog in Photoshop, the image dimensions may be 20.3 x 15.2 cm at 300 dpi. For example, we can now change the resolution to 72 dpi, but we have to make sure that the image is not converted; you should avoid that as a matter of principle. Ok, so we now enter 72 dpi - and we get an image size of 84.7 x 63.5 cm. So wonderful picture enlargement? Nonsense, the picture is still exactly the same as before. Not a single pixel has been changed, deleted, or added. As I said, numbers have no meaning for the pixel image. The cameras do enter secondary dimensions when recording, but the entries can vary from camera to camera, with the same pixel size, a sign of arbitrariness.

Another example, also easy to understand. We start again from our picture 2400 x 1800 and want to print it. With the printing we leave the image editing, and the printing is the only case where the metric size (in cm) is asked. The printer has to know how big the picture should be. Before we print, we want to give the image the desired print size, for example 14.5 x 10.9 cm (postcard size). So we open the size dialog in Photoshop and maybe find the dimensions 33.9 x 25.5 cm at 180 dpi displayed there. Now let's go and change the size to the postcard size. Then we print, and the printer promptly delivers the image exactly in the desired size. So are you all right? Seems so, and yet something possibly bad has happened. The next day we reload the picture, this time to make a larger print for the living room wall. We are horrified to find that the image only has 1028 x 771 pixels, which is definitely not enough for the larger print. Reason for the mishap: When entering the postcard format the day before, we allowed the image to be calculated smaller in terms of pixels. It was not noticed because we only paid attention to the size in cm. Now the picture can no longer be used for larger prints, it has been broken. It wasn't even necessary to determine the print size beforehand. Each printer driver allows the image to be scaled to the desired size on the spot, immediately before printing. This only happens temporarily, for each print. The image itself remains untouched.

Conclusion: Stay away from the relative resolution, stay away from the metric size in cm, at least in the editing program. There are exceptions, but they would only be confusing at this point. And of course the desired print size must be set in the printer driver before printing, either directly in cm or by dragging the image in the preview. If you want to determine the print size during image processing, which is not recommended, but possible, then you should do so with great caution. The image may not be recalculated. If this is observed, the program enters the desired print size in the image, but only as pure information for the printer.

How many pixels?

Yes, how many pixels should you set in the camera? Simple answer: As much as it is Use the image appropriately is. Let's start with the images to be printed. This is where the relative resolution (dpi) comes into play, but only as a guide. An optimal print (optimal = doesn't get better) is achieved with a resolution of around 300 dpi. Since we are not very familiar with the inches, let's do the conversion first: 300 pixels per inch means 300 / 2.54 = 118 pixels per cm. We can safely count on 100 pixels, because all of these are only guidelines. So we have a wonderful rule of thumb: 100 pixels per cm, that is 1000 pixels per 10 cm, 3000 pixels per 30 cm (A4, longer side) etc. 3000 x 2000 is the optimal number of pixels if in A4 format should be printed. Anyone who basically only 'orders' exposures or prints in the format 15 x 10 cm in the DM-Markt or from Rossmann's machines will be well served with 1600 x 1200 pixels. And if he took the photos with only 1200 x 800 pixels, there is hardly any loss of quality to be seen. The 300 dpi are not a law, just a guide.

Once again: The number of pixels is decisive. If a resolution of 96 dpi is given for a picture and someone thinks he could put things right by entering 300 dpi, then the good guy is wrong. Either the number of pixels remains unchanged, then the setting is nothing more than an exercise in entering numbers. Or the image is actually extrapolated. But nothing is gained thereby, because the substance of the picture cannot simply be increased.

The relationships are very clear if the images are intended for any application in the computer. It can be a slide show or some kind of presentation. However, this can also be used on a website. always is operated with pixel numbers, only with pixel numbers. Size specifications in cm have absolutely no place on the screen. And when someone speaks of screen-compatible resolutions (e.g. 72 or 96 dpi), then they are simply talking nonsense. Nor does the nonsense make sense because these numbers have persisted in some programs or in the minds of 'experts'.

So no resolutions, only numbers of pixels. You orientate yourself on common monitors or projectors. The screen is often already filled with an image width of 1000 pixels, and the displaying program is not always able to scale the images larger or smaller. So if the images are to stay in the computer from the outset, then it is a good idea to set something like 1600 x 1200 pixels in the camera, maybe only 1024 x 768. A pixel size of 800 x 600 is a bit tight these days.

Perhaps a small example: An amateur photographer, very quality-conscious, photographs some motifs with the great resolution of 5400 x 3780 pixels. Sure, according to the rule of thumb, he can make prints of up to 54 x 38 cm. But is that enough for a billboard measuring 2.00 x 1.50 m? he asks worried. One can only say: Dear photo friend, that's enough. People aren't going to crawl into the square with a magnifying glass and count pixels. Such images are viewed from a suitable distance and, from the required distance, appear similar to smaller images that are approached closer. The 300 dpi as a guide apply to the close viewing distance of perhaps 30 cm. At a distance of 3 m, 30 dpi is mathematically sufficient, which in turn derives the rule of thumb that around 10 pixels are required per cm. This is of course theoretical, because the picture should still reveal something, even if you get a little closer. You just have to watch the people in a museum.

And what about scanning?

In this post I cannot go into the specifics of the various scan programs. But the user cannot avoid becoming thoroughly familiar with the program. Two examples should make the problem clear:

Jakob took great pictures in the high mountains with an analog 6x7 and then scanned the negatives. There are still many who believe in the fairy tale of the good negative and the bad sensor. Anyway, Jakob sets a resolution of 150 dpi in the scan program because he has found that this is a reasonable resolution. It is, based on the final picture, but here the shot refers to the overall small negative. The results were beyond discussion. The good Jacob should have scanned with perhaps 1600 dpi in view of the small originals. The scanning program was quite confusing in this regard. Jakob has to start the scanner again. And scanning takes time!

In addition to mountain photos, Jakob has a second passion: he loves scanotypes. Small objects, flowers, etc. are placed on the scanner's glass plate and scanned. Looks great.After his experience with the medium format negatives, he sets the resolution to 4000 dpi. If so, he tells himself. But unfortunately he does not get to see the result, because that would be an image of more than 1 gigabyte. A normal PC just can't handle that, and let's be honest: 1 GB, isn't that a bit much for a daft carnation? And that takes, this time right, with a coffee break and a walk in between. The carnation was totally withered at the end of the scan.

It's two things I'm getting at. On the one hand, it is important to consider what type of template is. On the other hand, it depends on what the resolution refers to, whether on the template or on the desired result. This is not necessarily immediately apparent in the scan program; there are even programs that sometimes proceed this way, depending on whether the standard mode or the expert mode is switched on. The user looks in vain for clear indications. Therefore my urgent advice to explore the scan program carefully with various test scans.

In brief: Larger originals such as documents, papers, pictures, etc., which should normally be printed in 1: 1 size, can be scanned with a relatively low resolution. 80-100 dpi is sufficient for documents, photos can be scanned with 200-300 dpi. If a photo is then to be printed twice as large, you simply double the dpi number. Incidentally, this is the only case where entering the image size in cm makes sense. The scanner records the size of the original as the image size, and you don't need to worry when printing that the print is exactly the same size as the original.

Negatives, slides, etc. will always be shown enlarged later. It would be nice if you could simply specify a number of pixels here, as with the camera. But - who knows the strange thoughts of scanner manufacturers and their programmers? We have to do some math here. With a 35mm negative, the result should be a possible print size of 30 cm (long side). So 3000 pixels are needed for this (see above). 3000 pixels on 3.6 cm, that corresponds to 3000 / 3.6 * 2.54 = 2117 pixels on one inch, i.e. a good 2000 dpi. I will spare myself the calculation example with a medium format negative at this point. That means we don't really need to do the math. In relation to the pages, the medium format is roughly twice as large as a small picture. As a result, only half the resolution is required. With 1500 or 1600 dpi you can do very well to get the best prints up to A4 format.

However, these considerations only apply if the resolution relates to the original under the scanner lid. However, some scanning programs relate the resolution to the final image, however large it is desired. Is that easier? I don't know and prefer to rely on the calculator next to the scanner.


The tricky thing about the scanner settings is that you have to take the detour of the confusing and ultimately unimportant secondary dimensions so that the scanner driver can internally calculate the only significant size, namely the number of pixels. Which, in turn, often remains in the dark.

The whole beating is only necessary because programs simply cannot detach themselves from the completely superfluous and confusing secondary sizes (dimensions in cm, relative resolution in dpi). Everything would be so damn simple and outrageously transparent if you only operated with pixel numbers, like in the camera. But as long as even Photoshop sets a bad example, we can still hope for some interesting mistakes that make digital life not easier, but noticeably more colorful. Well, it doesn't have to be a gigabyte-sized carnation. Incidentally, the carnation photographer really does exist, of course his name is not Jakob. This scan friend described his problems with Gigabyte things in a specialist magazine for several pages and in the end, I think, bought a new computer. So that he can also process gigabyte roses and gigabyte handprints professionally and with high quality. Hm, consistent, but stupid. I can't help but think about the guy. Oh, I tried the scanotype process and got great results - with 40 dpi.